The name is the button.

An agent writes the code now, and a human thumb hovers over Accept all a hundred times a day. The honest question of this decade is what, exactly, you just accepted. So I read the diff.

Eight years as a backend engineer in Stockholm, back in Tokyo since 2023, and I still ship something small most nights, usually badly. My one rule: the only real review of a model is the thing you built with it. That is why there is no leaderboard on this site.

Around a hundred and forty sources come through — the labs' own posts, the papers, the tools, the runtimes and languages nobody is watching yet, the chip under the demo, and the newsletters people actually forward. What is moving gets counted so the week reads at a glance, and every few days I explain one mechanism in plain words.

A story here is a headline, a short excerpt, one image and the link to whoever wrote it. We never reproduce the post; the images belong to their authors. A post written in Japanese appears in Japanese, exactly as it was published, and the link goes to the original.

— Yuki Halvorsen, Tokyo

← Accept All Archive

llama.cpp releasesTools

b11543: vulkan: use 4 rows for NVIDIA MUL_MAT_ID MMVQ (#29274)

October 10

Keep the existing RDNA3/4 policy unchanged and update only rm_id to use 4 rows for NVIDIA except pre-Turing.

Chips & compute

Read at llama.cpp releases ↗

One letter a week, starting soon.

Latest

b11543: vulkan: use 4 rows for NVIDIA MUL_MAT_ID MMVQ (#29274)

Related

Toolsb11538 llama.cpp releases

AsiaShengShu Releases Vidu Q4 Preview With 4K Output and Launch Pricing From RMB 0.09 per Second Pandaily

AsiaSingapore data center operator DayOne explores $500m bond Tech in Asia

ValleyOxide Computer raises $445M to step up data center rack production SiliconANGLE

SiliconUpscale AI Partners With Nvidia While Competing With It The Next Platform

SiliconMaster AI Chip Principles With New IEEE Design Program IEEE Spectrum AI